When integrating advertisement and commercial content (generally referred to as “ads”) into search results, it may be useful to identify the general categories associated with the search query and provide ads associated with those categories. One approach is to determine the category directly from the results of the search query itself. However, with this approach, the search results must be known before the category can be determined for the purpose of ad integration. Another approach to find related content (e.g., ads) for a query is performed based on expanding the query based on historical search results shown with respect to past searches for the search terms or phrases within the query. This approach generates synonyms for various search terms and/or phrases when these search terms and phrases lead to search results that provide a statistical likeliness that the search term or phrase is synonymous with other term or phrases. This requires a high volume search query, such that previous search results exist for the query and can be analyzed to determine synonyms (e.g., including a category) accurately within an acceptable margin of error. Additionally, because the synonyms are driven from results across a large numbers of users and geographic areas, this approach does not offer geographic granularity (e.g., beyond country). Geographically specific generalization of a business to one or more categories, independent of the volume of past results, is useful in providing a wider array of content (e.g., including ads) to a user in response to a query.